Space-time analyses for forecasting future incident occurrence: a case study from Yosemite National Park using the presence and background learning algorithm

Researchers have developed a novel approach to predict the likelihood of wildland search and rescue (WiSAR) incidents in Yosemite National Park, leveraging machine learning and spatial statistics. Using the presence and background learning (PBL) algorithm, they generated monthly probability maps for 2011 that identified high-risk areas based on environmental variables such as distance to natural and anthropogenic features, vegetation, elevation, and slope, combined with historical incident data from 2001 to 2010. This method successfully forecasted the most probable locations and times for WiSAR incidents, providing actionable insights for resource allocation by decision-makers. The study not only addresses a critical real-world problem but also advances space-time analysis techniques, offering broader applications in epidemiology and ecological niche modeling.

Doherty, P. J., Guo, Q., Li, W., & Doke, J. (2014). Space-time analyses for forecasting future incident occurrence: a case study from Yosemite National Park using the presence and background learning algorithm. International Journal of Geographical Information Science, 28(5), 910–927. https://doi.org/10.1080/13658816.2014.890202